mirror of
https://github.com/saymrwulf/onnxruntime.git
synced 2026-07-14 18:12:05 +00:00
ce1bc9b49c
**Description**: Describe your changes. add rocm ep doc. Perview Github page : https://ytaous.github.io/onnxruntime/ Added items Home page - [ ] Update matrix on home page: https://ytaous.github.io/onnxruntime/ - [ ] Link "Installation Instruction Follow build instructions from here" to onnxruntime build with ROCm(inference) page.(This link now point to preview page, will change later) Installation - [ ] Update installation page: https://ytaous.github.io/onnxruntime/docs/install/#training-install-table-for-all-languages Build onnxruntime - [ ] Add onnxruntime build with ROCm(inference): https://ytaous.github.io/onnxruntime/docs/build/eps.html#amd-rocm - [ ] Update onnxruntime build with ROCm(training): https://ytaous.github.io/onnxruntime/docs/build/training.html#gpu--rocm ExecutionProvider - [ ] Add ROCm ExecutionProvider navitagation: https://ytaous.github.io/onnxruntime/docs/execution-providers/ - [ ] Add ROCm-ExecutionProvider page: https://ytaous.github.io/onnxruntime/docs/execution-providers/ROCm-ExecutionProvider.html Performance - [ ] Add enable_rocm_profiling introcduction: https://ytaous.github.io/onnxruntime/docs/performance/tune-performance.html#profiling-and-performance-report - [ ] Add graph-optimation for ROCm ep: https://ytaous.github.io/onnxruntime/docs/performance/graph-optimizations.html#extended-graph-optimizations Co-authored-by: Ethan Tao <ettao@microsoft.com@orttrainingdev7.d32nl1ml4oruzj4qz3bqlggovf.px.internal.cloudapp.net> Co-authored-by: ytaous <4484531+ytaous@users.noreply.github.com>
1.5 KiB
1.5 KiB
| title | description | parent | nav_order | redirect_from |
|---|---|---|---|---|
| ROCm (AMD) | Instructions to execute ONNX Runtime with the AMD ROCm execution provider | Execution Providers | 11 | /docs/reference/execution-providers/ROCm-ExecutionProvider |
ROCm Execution Provider
{: .no_toc }
The ROCm Execution Provider enables hardware accelerated computation on AMD ROCm-enabled GPUs.
Contents
{: .no_toc }
- TOC placeholder {:toc}
Install
Pre-built binaries of ONNX Runtime with ROCm EP are published for most language bindings. Please reference Install ORT.
Requirements
| ONNX Runtime | ROCm |
|---|---|
| main | 5.2.3 |
| 1.12 | 5.2.3 |
| 1.12 | 5.2 |
Build
For build instructions, please see the BUILD page.
Usage
C/C++
Ort::Env env = Ort::Env{ORT_LOGGING_LEVEL_ERROR, "Default"};
Ort::SessionOptions so;
int device_id = 0;
Ort::ThrowOnError(OrtSessionOptionsAppendExecutionProvider_ROCm(so, device_id));
The C API details are here.
Python
Python APIs details are here.
Performance Tuning
For performance tuning, please see guidance on this page: ONNX Runtime Perf Tuning
Samples
Python
import onnxruntime as ort
model_path = '<path to model>'
providers = [
'ROCmExecutionProvider',
'CPUExecutionProvider',
]
session = ort.InferenceSession(model_path, providers=providers)